How Many People Visit a Website? These 6 Free Tools (With Paid Features) Can Help You Analyze That

Checking how much traffic a website is getting is crucial when you want to analyze a competitor or track your own site so you can make adjustments accordingly. So, to check monthly/weekly website traffic, you need different tools to keep track of that data. SEO, information researchers, and digital marketing experts use different traffic checker tools every day. Although some of them don't show the exact number of traffic a website is receiving, they give solid estimates and reveal which keywords are helping in driving the traffic towards the website.

These website traffic checker tools are important if you want to do content planning, keyword research, and competitor analysis. This article is for you if you are looking for free website checker tools (that also offered better paid features), as we will look at what insights you can get from them and how to use them to track your (or any other) website’s traffic.

Why is Checking Website Traffic Important?

Knowing how much traffic a website gets, especially your competitors, is important as it can give you insights about how to improve your own marketing strategy to get maximum eye-balls.

When you have the right traffic checker tools, you can:

• Get to know a lot of top traffic sources and know which channels drive the most visitors to the website.
• You can discover the keywords' performance of the website and see which keywords are helping the competitor’s website rank and which ones they are losing.
• You can find out which pages and posts bring high traffic through traffic checker tools.

• You can also get to know which ads the competitors are running and where.

Website traffic checker tools aren't used for copying the competitor’s content; they are used to spot your own website’s content gaps and what changes you can make to create more useful and better resources that attract the same audience as your competitor. This way, you can get organic traffic and can add things that your competitors have missed.

It is also important to remember that tracking website traffic isn't the whole thing. You have to check the digital presence of your competitors like email campaigns, social media, and ads, to get the whole picture about their digital strategy.

Top 6 Best Website Traffic Checkers

Here are the top 6 best website traffic checkers that can help you know the stats about websites:

1. SimilarWeb
2. Semrush
3. Ahrefs
4. NeilPatel's Website Traffic Checker
5. Sitechecker
6. SERanking

Now let’s take a look at them in detail.

1. SimilarWeb

What makes SimilarWeb different from others is its Chrome extension. Once you install it, you can click on the extension to see basic traffic metrics of any website and check again for more detailed data. It has many useful tools for marketing and SEO professionals.

Features:

• You can check the global ranking of a website on SimilarWeb.
• It can also help you find the top traffic countries of the website.
• You can also find the visits over time on the website.
• You can check traffic sources as well as keyword rankings.
• SimilarWeb also gives you organic traffic estimates.

If you are just starting out, the free version of SimilarWeb is enough, but you can unlock more features in the premium version. There is also a full suite of tools for marketing professionals on SimilarWeb. The Competitive Research tool helps in comparing market share with selected competitors, and you can access many of the same metrics as Semrush and Ahrefs.

• Also read: Top SEO Tools for 2025 Every Marketer Needs to Dominate Google Search Results

We chose SimilarWeb as our number one pick because, during our research, we were specifically looking to find out how many daily visitors/users ChatGPT receives, a figure not publicly disclosed by OpenAI. With no official numbers available, we turned to SimilarWeb for a reliable estimate (combined from both mobile and desktop devices). Among all the tools we tested, SimilarWeb provided the most comprehensive and consistent traffic data, helping us determine that ChatGPT attracts approximately 170 million daily visitors. (You can read the full post here, if you want.)

2. Semrush

Semrush is undoubtedly one of the best website checker tools, and it does a lot more than that. It has a lot of features that can help you grow your traffic and improve your SEO.

Features:

• Semrush finds out which keywords competitors are ranking for.
• You can get an idea about how much traffic your competitors are getting.
• You can track keyword rankings from your own website.
• It can also mine your competitors' backlinks to find opportunities for your website.

You just have to give Semrush the URL of the website and it can give you information about top organic keywords, traffic, backlinks, top landing pages and more. The Domain Overview report by Semrush helps you take a look at your competitor’s website, and there's also another feature called Traffic Analytics that shows the total number of visits, traffic sources and countries that send traffic to the website.

There is also another tool known as the Keyword Magic Tool, which helps you find the best keywords for your website. All you need to do is type in a keyword and Keyword Magic Tool gives you related keywords, search trend, search volume, keyword difficulty and other keyword-related things. You can use the limited version of Semrush to get an idea about its website traffic checker, as its paid version may seem too costly for some.

3. Ahrefs

Another nifty traffic checker tool is Ahrefs, which gives you important insights about what your competitors’ websites are up to. There is also a full suite of SEO tools that go beyond competitor research.

• Related: 2025’s Best Social Sentiment Tools to Help Brands Know Customers Better

Features:

• Ahrefs can help you measure a competitor’s organic traffic.
• You can see estimated monthly visits on the website.
• Ahrefs can help you view organic keywords on the website.
• It can analyze traffic value and competing domains.
• It also performs technical SEO audits.
• Ahrefs also finds content gaps in your website.

All you need to do is paste the website URL and you will get all the information about the competitor's website. You can also use Ahrefs to find broken links. Ahrefs also has an Organic Keywords report which helps you discover the search terms that others rank for and do keyword research to optimize your own page.

4. NeilPatel's Website Traffic Checker

Website Traffic Checker is a free and simple tool for checking website traffic. It has many useful features and is part of Neil Patel’s Ubersuggest keyword and SEO platform, which makes it a non-standalone tool. One draw back is though it shows too many popups whenever you try to visit the tool.

Features:

• It gives you estimates for organic traffic on websites.
• You can check organic keywords on websites using Website Traffic Checker.
• You can also take a look at the top pages of the website.
• You will also get estimates for domain authority and backlinks on Website Traffic Checker.

On the free version of the Website Checker tool, you get data from the past 4 months. If you want to check data prior to that, you have to upgrade to the paid version of Ubersuggest.

5. Sitechecker

Sitechecker also has a Website Traffic Checker tool which lets you see organic website traffic for free but in a limited way.

Features:

• Sitechecker gives you three months of website traffic overview
• You get to know about the engagement metrics of the website.
• Sitechecker gives you metrics about traffic by country and source.
• It also tells you the top five organic keywords of the traffic.
• You can also find out the top referrals and competitors of the website using Sitechecker’s website traffic checker.

In its free version, you get a basic understanding of a competitor’s website traffic. For more detailed data, you have to subscribe to its paid plan.

6. SERanking

SERanking’s Competitor Analysis Tool is part of its full SEO and marketing dashboard. When you give it a website URL, you get different tools and insights about your competitor’s website.

Features:

• By using SERanking, you get to see domain and page trust.
• It gives you insights about estimated organic traffic, organic keywords, and organic competitors.
• You also get to know about total keywords, and referring domains and backlinks.
• SERanking also tells about SERP features of the website.

SERanking’s Competitor Analysis Tool is a very useful tool that gives you a decent amount of data in the free version. If you want to track your own keywords or want more information, you have to buy its monthly plan.

Frequently Asked Questions

1. How can I see the traffic of the website?

You can use different tools like Google Analytics, for the websites/blogs you own, while Semrush, and SimilarWeb to see the traffic of any other website (or the one you owned) and get insights about user behavior, page views and web performance.

2. Can I see how much traffic a website gets for free?

Yes, you can see different website traffic for free using limited versions of tools like SimilarWeb, Semrush or Ahrefs.

3. Is there a free tool or software to get info about a website’s visits traffic?

Yes, the tools mentioned in this post can help you estimate a website’s monthly traffic. While you can divide this number to get rough weekly or daily figures, keep in mind that traffic often fluctuates, so such breakdowns may not be precise. Accurate traffic numbers are usually only available to the site owner via tools like Google Analytics or Cloudflare Analytics.

Read next: 

• These Are the Best AI Video Generators for Creating Stunning Content in Minutes

by Arooj Ahmed via Digital Information World

Brute Force-Friendly Loophole in Google Recovery Flow Got Fixed After Researcher Demonstrated Full Number Exposure

A now-patched vulnerability in Google's account recovery system briefly opened the door to a subtle but powerful form of data exposure, allowing determined attackers to retrieve a user's private recovery phone number without their knowledge. The weakness, while no longer active, once relied on a series of overlooked interactions between separate account features rather than a single critical flaw—a method that made it difficult to detect and even harder to trace.

The issue came to light after an independent researcher, operating under the name Brutecat, noticed that one of Google's recovery forms continued to function even when JavaScript was disabled. This small detail, seemingly harmless at first glance, proved to be the foundation of a broader attack chain. By using this non-scripted version of the form, it became possible to query whether a given phone number or email address was linked to a specific Google display name, a process that unintentionally provided attackers with confirmation about the validity of recovery methods tied to individual accounts.



Once that connection was established, the next step involved the standard password reset process, which displayed a partially hidden version of the associated phone number. While the masked format protected most of the digits, enough information remained visible to guide a targeted brute-force approach—particularly when the attacker already knew the victim's country of residence. With the international dialing code fixed, the number of unknown digits dropped to as few as six or seven, a range that could be cracked in a matter of seconds by running automated scripts on relatively modest hardware.

To improve the accuracy of the attack, the researcher also found a creative way to extract a user's full name by exploiting a function within Looker Studio, Google's data visualization platform. By creating a shared document and transferring ownership to the target account, it triggered the appearance of the user’s complete display name in the interface. This detail, when combined with the earlier methods, allowed the attacker to craft highly specific requests that reduced guesswork and accelerated the brute-force phase of the exploit.

Testing revealed that the process was not only effective but alarmingly fast. With a simple server setup, the researcher demonstrated that thousands of number combinations could be tested every second. In some countries, the full recovery number could be exposed in under half a minute. For others, it took just a few minutes longer—still well within the range of practical exploitation.

Although the underlying flaw has since been addressed, with Google confirming the deactivation of the vulnerable endpoint and implementation of additional safeguards, the incident underscores how separate components of a digital ecosystem can be manipulated in tandem to bypass intended protections. While no known widespread abuse occurred before the fix was deployed, the method serves as a reminder that security is not only about strong individual barriers but also about how those barriers connect.

Read next: Apple Overhauls Software Aesthetics, Launches AI Tools and Developer Upgrades in Privacy-Centered WWDC 2025 Refresh
by Irfan Ahmad via Digital Information World

Apple Overhauls Software Aesthetics, Launches AI Tools and Developer Upgrades in Privacy-Centered WWDC 2025 Refresh

Apple has taken a deliberate step toward a more cohesive and intelligent future for its devices, using this year’s Worldwide Developers Conference to unveil what may be its most far-reaching software refresh in more than a decade. At the heart of these changes lies a complete redesign of the interface language across its platforms, which the company now refers to as Liquid Glass—a visual style that reflects light, adapts to user interaction, and adds depth to previously flat screen elements. This new aesthetic, which replaces the static transparency of older designs with a more responsive and immersive surface, has been engineered to appear across iPhones, Macs, iPads, and other devices, pulling the ecosystem into tighter visual alignment.

In parallel with the design overhaul, Apple introduced a new naming system for its operating systems. Rather than assigning each platform a distinct version number, the company will now identify releases by year, simplifying the naming convention that had grown inconsistent over time. All major platforms, from iOS and iPadOS to macOS, tvOS, and watchOS, will now adopt labels like 26, creating a unified software identity under a single annual release cycle.

Much of the company’s attention this year turned toward the integration of artificial intelligence, though Apple’s approach remains cautious and grounded in privacy. Under the umbrella of Apple Intelligence, a collection of new features has been built to operate primarily on-device. Live Translate, for example, enables real-time conversation translation directly within apps like Messages and FaceTime, without requiring data to be sent off-device. Another addition, known as Visual Intelligence, recognizes content displayed on screen and suggests actions accordingly, such as adding an event to the calendar if it detects date and time details within a viewed image or webpage.

Developers will be given access to Apple’s foundational AI models through a new framework, allowing them to embed machine learning into their apps while maintaining user privacy and offline functionality. Within this shift, Apple also opened the door to integration with third-party models such as OpenAI’s ChatGPT, which can now be used within certain apps like Image Playground. However, the company emphasized that no data will be shared externally unless a user gives explicit permission, reinforcing its privacy-first stance in contrast to cloud-reliant competitors.

Software tools have also seen major upgrades, particularly for developers. Xcode now includes built-in support for AI-assisted programming, offering suggestions, debugging help, and code testing with support for external models as well. This move appears designed to lower the barrier for building apps that incorporate Apple Intelligence, even for smaller development teams.

Beyond the system-wide updates, a range of feature-focused additions have been announced. A new dedicated Games app will act as a central hub for players, bringing together installed titles, achievements, leaderboards, and a social layer that enables users to see what friends are playing and initiate multiplayer sessions. The app introduces score-based challenges and makes it easier to manage and discover gaming content across Apple platforms.

Other systems are receiving their own tailored improvements. macOS 26, nicknamed Tahoe, brings AI enhancements to Shortcuts, enabling faster automation and task execution. On iPad, new features such as Preview and markup tools offer users greater flexibility for editing and exporting visual content, particularly when using Apple Pencil. The Apple Watch gains new gesture-based controls and a Workout Buddy function that uses on-device intelligence to track fitness in a more personalized way. Apple TV now supports profile switching upon wake, a karaoke feature, and a refreshed Liquid Glass layout that simplifies navigation. VisionOS, designed for Apple’s AR/VR headset, will support additional accessories, including Logitech’s Muse stylus and Sony’s VR2 controller, expanding the input options for immersive experiences.

Even Apple’s peripheral services saw meaningful updates. AirPods will now support higher-quality audio recording and offer remote camera control, while CarPlay users gain message pinning, live activities, and interactive widgets. Apple Wallet introduces new travel features, Maps learns regular commute patterns, and iMessage adds polling tools and enhanced customization for chats.

While the presentation lacked the theatrics that often accompany major hardware reveals, Apple’s software direction this year carries significant weight. Instead of relying on spectacle, the company has chosen to embed progress more subtly into the everyday user experience, combining design, intelligence, and cross-platform coherence in a way that suggests quiet confidence rather than urgency. In doing so, Apple may not be trying to outpace its competitors on AI — but it is clearly intent on ensuring every piece of its ecosystem moves forward in lockstep.

Read next: Where in the World Are LinkedIn Users Most Likely to Call Themselves CEOs?
by Irfan Ahmad via Digital Information World

Where in the World Are LinkedIn Users Most Likely to Call Themselves CEOs?

• One in 18 LinkedIn users in Switzerland lists “CEO” as their title. In Switzerland, 5.48% of all LinkedIn users list “CEO” in their profile, the highest share recorded in the entire analysis.
• One in every 23 LinkedIn users holds the CEO title in Sweden, a rate that places it well ahead of countries like Italy and France.
• Germany has the highest total number of CEOs on LinkedIn.

A recent data review has revealed surprising concentrations of self-declared CEOs across LinkedIn, especially in smaller European economies. The findings follow a platform-related incident where a user momentarily assumed the identity of LinkedIn’s chief without authentication, sparking questions around professional identity standards.

The study, curated by Heepsy, focused solely on countries with over a million LinkedIn users. The aim was to draw meaningful comparisons by tracking the share of members who label themselves as chief executive officers.

Topping the chart, Switzerland displayed the densest clustering. Roughly one out of every 18 profiles there carries the CEO tag. This equates to 74,000 individuals among its 1.35 million user base. The high frequency may signal a strong startup culture or liberal title usage in business circles.

Just behind, Sweden registered similar behavior, with about 71,000 professionals—roughly one in every 23—tagged under the executive tier. Spain stood next in line, where more than 150,000 users listed themselves as company heads, comprising just over 4% of its total LinkedIn crowd.

Germany, despite topping the chart in volume, held the fourth spot by percentage. Out of 4.42 million LinkedIn profiles there, 185,000 identify as company leaders. That makes the ratio slightly less intense than in neighboring countries.

Denmark, with its comparatively modest LinkedIn population, saw around 43,000 listings for chief executive roles, earning it a spot in the top five. The frequency—just under 4%—reflects a high saturation of leadership identifiers.

Poland followed with a 3% share, translating to over 50,000 users labeling themselves as CEOs. While the total isn’t among the highest, the relative density suggests an active trend of status signaling or title normalization among digital professionals.

Further down the list, Belgium saw under 3% of its members opting for the CEO descriptor. Though the proportion is smaller, it still outranks several larger nations. Italy, despite having four million users, showed a lower percentage—less than 3% of its profiles declared a chief executive role.

The Netherlands, often noted for egalitarian corporate norms, posted a comparatively reserved rate. Roughly 2.26% of profiles used the CEO tag, pointing to a culture less inclined to formal executive labels.

France, despite leading the group in total LinkedIn users, trailed at the bottom. Only 1.76% of French professionals claimed the CEO title—roughly one in 57—marking it the most conservative among the ten nations surveyed.

"LinkedIn has evolved from just a job board to a key platform for showing off our professional identities. The way we present ourselves, particularly through our job titles, reflects interesting cultural views on leadership and success. The title of CEO now focuses more on how we wish to be seen by others rather than just our role in a company. These cultural variations in personal branding are very important as companies work globally and base hiring choices on online impressions.", Tabi Vicuña, Founder of Heepsy, commented on the study.

Also read: Study Warns of Surge in Security Breaches Linked to Social Media Impostors

Outside Europe, the trend continues with wide disparities. The United States, home to the world’s largest LinkedIn population at 236 million, recorded just over 2.4 million self-labeled CEOs—around 1.03%. Canada followed a similar pattern, with 180,000 out of 26 million users claiming the executive tag, marking a modest 0.69% share. In contrast, South Korea posted a higher percentage at 1.66%, translating to 72,000 CEO profiles from a pool of 4.3 million. Nigeria stood out among African nations, with 146,000 CEO claims across 10.3 million accounts—about 1.42%—surpassing even Japan’s 1.24% figure.

Elsewhere, emerging economies displayed varying levels of title adoption. Ghana logged a 1.13% CEO density, while Kenya (0.61%), South Africa (0.55%), and Egypt (0.32%) showed more tempered behavior. Southeast Asia reflected cautious usage overall. Vietnam, Malaysia, and Indonesia all reported CEO listings below 0.4%, despite collectively hosting tens of millions of users. India, despite being the second-largest LinkedIn market with over 135 million members, registered only 415,000 CEO-tagged profiles—amounting to 0.31%.

Latin America showed similar restraint. Brazil, with 71 million users, saw just 283,000 CEO entries, or 0.40%. Mexico, Colombia, and Argentina each fell below the 0.4% mark as well. European outliers like the United Kingdom (0.65%), the Netherlands (2.26%), and Italy (2.68%) offer contrast to places such as Qatar (0.34%), Jordan (0.31%), and Turkey (0.29%), where the label remains less saturated.

Taken together, the figures reflect more than just professional labels—they expose how digital economies shape perception, branding, and hierarchy. In regions with a growing startup presence or weaker gatekeeping of executive roles, the CEO title may serve more as a positioning tactic than a reflection of organizational structure. As professional networks expand their global reach, such variations raise questions about the consistency and meaning of executive identity in the online space.

Country	LinkedIn Users 2024	LinkedIn CEO Profiles	% of CEO Profiles
Switzerland	1,350,000	74,000	5.48%
Sweden	1,600,000	71,000	4.44%
Spain	3,550,000	155,000	4.37%
Germany	4,420,000	185,000	4.19%
Denmark	1,110,000	43,000	3.87%
Poland	1,696,000	52,000	3.07%
Belgium	1,280,000	38,000	2.97%
Italy	4,000,000	107,000	2.68%
Netherlands	2,870,000	65,000	2.26%
France	8,600,000	151,000	1.76%
South Korea	4,338,000	72,000	1.66%
Nigeria	10,310,000	146,000	1.42%
Japan	4,500,000	56,000	1.24%
Ghana	2,840,000	32,000	1.13%
United States	236,000,000	2,420,000	1.03%
United Arab Emirates	8,660,000	77,000	0.89%
Angola	1,009,600	8,600	0.85%
Australia	15,950,000	125,000	0.78%
Cameroon	1,242,000	8,700	0.70%
Uganda	1,438,800	10,000	0.70%
Serbia	1,440,000	10,000	0.69%
Canada	26,000,000	180,000	0.69%
Hong Kong	3,516,000	23,000	0.65%
United Kingdom	42,700,000	277,000	0.65%
New Zealand	2,940,000	19,000	0.65%
Kenya	4,925,000	30,000	0.61%
Singapore	4,600,000	27,000	0.59%
South Africa	13,910,000	77,000	0.55%
Tanzania	1,362,000	7,400	0.54%
Lebanon	1,251,000	6,400	0.51%
Taiwan	3,576,000	18,000	0.50%
Puerto Rico	1,074,000	5,100	0.47%
Bangladesh	8,865,000	42,000	0.47%
Dominican Republic	1,925,000	8,600	0.45%
Oman	1,016,000	4,400	0.43%
Panama	1,344,000	5,800	0.43%
Kuwait	1,145,000	4,900	0.43%
Vietnam	8,297,000	35,000	0.42%
Sri Lanka	2,348,000	9,900	0.42%
Brazil	71,100,000	283,000	0.40%
Senegal	1,176,000	4,600	0.39%
Kazakhstan	1,561,000	6,100	0.39%
Ethiopia	1,179,900	4,600	0.39%
Nepal	1,713,000	6,500	0.38%
Mexico	22,780,000	85,000	0.37%
Thailand	5,464,000	20,000	0.37%
Malaysia	8,500,000	31,000	0.36%
Saudi Arabia	9,930,000	36,000	0.36%
Qatar	1,548,000	5,200	0.34%
Egypt	11,430,000	37,000	0.32%
Chile	8,680,000	28,000	0.32%
Jordan	1,800,000	5,600	0.31%
Uruguay	1,482,000	4,600	0.31%
India	135,400,000	415,000	0.31%
Turkey	16,380,000	48,000	0.29%
Costa Rica	1,791,000	5,200	0.29%
Guatemala	1,640,000	4,700	0.29%
Colombia	14,770,000	42,000	0.28%
Argentina	14,260,000	39,000	0.27%
Tunisia	2,213,000	5,800	0.26%
Ecuador	4,856,000	12,000	0.25%
Bolivia	1,696,000	3,800	0.22%
Indonesia	27,950,000	60,000	0.21%
Morocco	5,321,000	11,000	0.21%
Peru	10,040,000	20,000	0.20%
Venezuela	5,220,000	10,000	0.19%
Iraq	2,119,000	3,800	0.18%
Philippines	17,680,000	29,000	0.16%
Algeria	4,415,000	3,700	0.08%

Methodology:

To ground these comparisons in clarity, the analysis relied on a straightforward methodology. Only countries with at least one million LinkedIn users were included, filtering for platforms with sufficient traction to offer meaningful data. For each country, researchers examined three primary inputs, that is, the total number of LinkedIn users, the count of those who list “CEO” within their job titles, and the percentage this group represents out of the total user base.

This percentage, not the absolute count, determined the rankings. By focusing on proportional representation, the study aimed to uncover not just where executives are found in volume, but where the title appears most densely. The intent was to capture regions where leadership is most prominently self-declared, suggesting localized trends in entrepreneurial culture, title inflation, or digital brand positioning.

All figures were drawn from publicly available 2024 LinkedIn data and reviewed to ensure consistency across markets. The final dataset includes only those nations with verified, complete information—ensuring that the leaderboard reflects not popularity alone, but patterns of perceived authority within the global professional space.

Read next: Apple Study Questions AI Reasoning Models in Stark New Report
by Irfan Ahmad via Digital Information World

Apple Study Questions AI Reasoning Models in Stark New Report

Apple has thrown a spanner into one of artificial intelligence’s most hyped developments. A new internal study suggests that so-called "reasoning" models—AI systems designed to emulate step-by-step human thought—are far less capable than the industry might have hoped.

Published under the title The Illusion of Thinking, the paper from Apple’s machine learning team provides a comprehensive critique of reasoning-enhanced large language models (LLMs). These are versions of AI that attempt to solve problems by simulating multi-step logic. But when put through their paces, the study suggests, these models often collapse under pressure.

Thinking Isn’t Always Smarter

Apple’s researchers tested various models on puzzle-based tasks, where the difficulty could be precisely increased. The puzzles were chosen because they demand logical thinking and can’t be solved simply by guessing what “looks right.”

Performance was grouped into three categories:

• Simple Tasks: Basic models, without added reasoning chains, often performed better. They solved easier problems faster and more accurately.

• Intermediate Tasks: This is the sweet spot for reasoning models. They showed some improvement over regular LLMs—but only briefly.

• Complex Tasks: Once tasks became genuinely difficult, all models failed. Crucially, the ones built to reason did not just struggle—they regressed, producing weaker output and fewer steps of thought.

What stood out most was the unexpected drop in performance as complexity increased. Rather than ramping up their reasoning to meet the challenge, models began to offer less reasoning altogether. This suggests that their thinking ability may be more cosmetic than structural.

A Closer Look at the Cracks

In one test, models were asked to solve the Tower of Hanoi puzzle—an old logic game familiar to anyone who’s studied basic computer science. With just a few discs, most models performed fine. But as the number of discs increased, success rates plunged. Even when fed the exact algorithm, models couldn’t reliably carry it out.

The issue wasn’t memory or compute limits. These AI systems had room to process more steps. They simply didn’t. In many cases, they stopped reasoning halfway through, abandoned logic midstream, or returned to earlier incorrect ideas. Even a slight change in the way a question was worded could break their performance.

This behaviour hints at a deeper limitation. Rather than working through a problem logically, as humans might, these models appear to rely on surface-level pattern matching. The reasoning traces they output may look convincing—but Apple’s paper found they often have little to do with how the final answer is reached.

Models That Overthink and Underdeliver

One striking insight was how often models veered off course after finding a correct solution. Instead of stopping there, they kept going—sometimes reversing their own correct steps. This phenomenon, known as "overthinking," resulted in answers being buried under layers of unnecessary logic.

Researchers noted a sharp divide depending on problem complexity. In easier puzzles, models often solved the task early, then talked themselves out of it. In mid-range puzzles, they wandered through trial and error before arriving at a partial answer. On the hardest tasks, correct solutions vanished entirely.

In short, the more you ask these systems to think, the less helpful their thinking becomes.

Not a Death Knell—But a Warning Bell

Apple is not calling for the end of reasoning models. Instead, the report urges a rethink about what these systems actually do. It also warns developers and businesses not to assume that “more thought” from a model automatically equals better results.

“Reasoning models aren’t useless,” one senior researcher said, “but their strengths are very specific—and their weaknesses are too often overlooked.”

This message lands at a sensitive moment. As rivals like Google, Microsoft and OpenAI rush to build general-purpose AI systems that reason, plan and even argue, Apple is taking a different path. Its AI focus remains grounded in efficient, privacy-first features that run on users’ devices.

The company’s new Apple Intelligence tools—rolled into iOS 18 and macOS Sequoia—emphasise helpfulness over generality. Features like writing suggestions, notification summaries and image generation aim to make day-to-day tasks smoother, not to mimic human reasoning.

Between Promise and Premature Praise

While some in the AI community have welcomed Apple’s study as a much-needed reality check, others see it as overly cautious. Critics point out that reasoning models are still in their infancy and argue that dismissing their progress so early could limit innovation.

But Apple’s data is hard to ignore. The company’s researchers did not rely on abstract benchmarks or vague tasks. They built controlled environments, adjusted puzzle complexity step by step, and measured not only answers but the logic behind them.

The result is a study that asks one key question: if today’s AI still stumbles on problems that children—or even simple algorithms—can solve, how close are we really to machines that think?

For now, it seems, the illusion of reasoning may be just that—an illusion.

Image: DIW-Aigen

Read next: Tracking the World’s Clicks: Daily Rankings of Top Search Platforms
by Irfan Ahmad via Digital Information World

Tracking the World’s Clicks: Daily Rankings of Top Search Platforms

Google continues to dominate the global search landscape, handling an estimated 13.7 billion searches per day, according to new figures shared by Neil Patel Digital. This puts its share at approximately 26.97% of total daily global search activity across major platforms. Despite growing competition, Google remains far ahead of any individual rival, underscoring its central role in how people access information online.

In second place, and by some distance, is Instagram. The social media platform now processes 6.5 billion searches daily, reflecting how users are increasingly turning to visual platforms to discover trends, products, and content creators.

China’s Baidu comes next with 5 billion, followed by Snapchat with 4 billion. These numbers show how different regions and use cases continue to shape search behaviour globally. Amazon, as expected, sees high search traffic as well—3.5 billion queries per day, driven by product lookups and shopping intent.

• Related: Traditional Search Holds Firm At 10% Use; AI Lags While Zero-clicks And Internal Google Clicks Rise

Other major platforms include YouTube with 3.3 billion, and LinkedIn, which surprisingly follows closely at 3.2 billion. Users are clearly searching for both educational and professional content at scale.

Pinterest and the Google Play Store register 2.4 and 2.1 billion daily searches, respectively. While not at the very top, both maintain steady engagement, especially among mobile users and niche audiences.

Further down, Facebook still draws 1.5 billion daily searches, and Yahoo, though past its prime, sees 1.1 billion. TikTok and ChatGPT each report around 1 billion, a figure that highlights how content discovery and conversational AI are beginning to play a more central role in how people search for and process information.

Reddit logs 900 million daily searches, primarily from users seeking real discussions or niche answers. Bing sits at 600 million, while both X (formerly Twitter) and Apple’s App Store close out the list with 500 million each.

Platform	Daily Searches
Google	13.7 billion
Instagram	6.5 billion
Baidu	5.0 billion
Snap	4.0 billion
Amazon	3.5 billion
YouTube	3.3 billion
LinkedIn	3.2 billion
Pinterest	2.4 billion
Google Play Store	2.1 billion
Facebook	1.5 billion
Yahoo	1.1 billion
TikTok	1.0 billion
ChatGPT	1.0 billion
Reddit	0.9 billion
Bing	0.6 billion
X	0.5 billion
Apple App Store	0.5 billion

The data reflects more than just user preferences—it signals a shift in what "search" means. It’s no longer just about finding websites. Increasingly, people search through images, conversations, short videos, and shopping apps, depending on their goal.

Google may still be far ahead, but the way people find information is clearly evolving. Platforms that integrate search naturally into their experience—whether social, commercial, or creative—are beginning to define the next phase of digital discovery.

Read next: AI Search Still Sends Most Traffic from Desktops, Not Mobiles
by Irfan Ahmad via Digital Information World

AI Search Still Sends Most Traffic from Desktops, Not Mobiles

Desktop devices continue to dominate referral traffic from AI-powered search engines, according to new data from BrightEdge. The research examined how users engage with leading AI search tools and found that most website visits driven by these platforms originate from desktop devices.

Referral Traffic Skewed Toward Desktop

BrightEdge analysed traffic patterns across key AI search engines, including ChatGPT, Bing Copilot, Google Gemini, and PerplexityAI. Despite the increasing popularity of mobile browsing, these platforms direct the vast majority of their outbound traffic from desktop users.

The breakdown is as follows:

• Perplexity.ai: 96.5% desktop, 3.4% mobile
• ChatGPT: 94% desktop, 6% mobile
• Bing: 94.4% desktop, 4.5% mobile
• Google Gemini: 91% desktop, 5% mobile
• Google Search: 44% desktop, 53% mobile, 2% tablet

Google Search is the only major AI platform where mobile referral traffic exceeds that of desktop.

Mobile Bottlenecks Affect Referral Behaviour

One explanation for the disparity is how AI apps operate on mobile. For example, the ChatGPT mobile app displays in-app previews when a user taps a link. This requires a second tap to open the original source, which may reduce the number of users reaching external sites. On desktop, by contrast, links lead directly to the target website.

Similar patterns are observed with Perplexity and Bing, where in-app design choices appear to limit mobile outbound traffic.

Google Benefits from Browser Defaults

Google’s higher mobile referral rate is likely influenced by its position as the default search engine on many mobile browsers, particularly Safari. According to the data, 58% of Google’s mobile search traffic to branded websites in the US and Europe originates from iPhones. With Safari installed on nearly one billion devices, this default setting has a significant impact on mobile traffic flow.

Search Marketing Implications

The data suggests that desktop remains the primary source of AI-driven website visits, even as mobile usage dominates general web traffic. For businesses tracking search referral performance, AI traffic is still largely shaped by desktop engagement and platform integration with browsers.

Unless AI tools find ways to improve outbound traffic handling on mobile, especially within app environments, this trend is expected to continue.

Read next: Deleted ChatGPT Conversations Weren’t Really Deleted — And Now OpenAI Is Pushing for ‘AI Privilege’
by Irfan Ahmad via Digital Information World